INDEX

Explanations

terms related to power and wattage specifications

oai_token-act-pair · gpt-4o-mini Triggered by @bot

electrical power measurements

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.20.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 DialogInterface

-0.41

IsMutable

-0.38

CGContext

-0.38

 szczegó

-0.38

GraphicsUnit

-0.37

➟

-0.36

 testigos

-0.36

uties

-0.36

 Carrillo

-0.36

 understandable

-0.36

POSITIVE LOGITS

 wattage

0.68

 power

0.66

Power

0.65

 Power

0.64

power

0.63

Datuak

0.60

 ſch

0.60

kW

0.60

功率

0.59

 Perſ

0.59

Activations Density 0.026%

terms related to power and wattage specifications

electrical power measurements

No Comments

No Known Activations